START
Chapter 5: Recommendations and Conclusion
5.2 Predicted Performance when HVAC Power Lines are Uprated for HVDC Application
From the research study, we can draw the following predicted conclusions when HYAC power lines are converted and up rated for HYDC application. The historical performance of the HYAC lines can be sourced from the utilities performance reports [43].
The predicted performance under HVDC operation will be as follows:
I. TV Interference - This will not be an issue as ion migration for distances beyond the
2. Positive Corona - this will contribute to radio interference; again at the edge of the servitude and beyond the levels calculated are very low and will be acceptable.
3. Audible Noise - again the contributor will be positive corona; this will be acceptable along the line route; higher levels could be at the converter station and will emanate from the converter transformers.
4. Pole to ground faults for the case of servitude fires will occur at mid-span; protection and converter control to be designed to manage these faults; self healing using voltage management could be a strategy. An outcome will be enhanced quality of supply as compared to AC systems.
5. Given the high electric fields in the air gap of the tower windows, it is expected that bird streamers will cause electrical faults. The installation of bird guards IS
recommended and should be included in the scope of the re-insulation proposal.
6. Silicone rubber insulators will deliver the higher creepage for the given connecting length; 30mm/kV affordable and is required. The hybrid toughened glass for DC application and coated with silicone rubber will make an ideal design for external insulation.
Silicone rubber technology for external insulation is now an acceptable and recommended insulator for external insulation of DC systems. The technology has matured. Work done on silicone rubber insulators corona testing [44] and artificial contamination [45] testing under laboratory conditions show that the new DC silicone rubber operates well within the specified limits and was recommended for the uprating of the Pacific DC Intertie. To date, the installed insulators are performing well. Back at Eskom, a few samples of these insulators are operating with no incidents. In the case of radio interference tests, a 40 dB RIV level was recorded for the positive 550 kV pole voltage whilst a 32 dB RIV level was recorded for the negative 550 kV pole voltage.
Visual corona extinction was achieved at +555 kV for the positive pole whilst -565 kV was recorded for the negative pole. In the case of the contamination tests, no flash over occurred for three consecutive tests. Here 515 kV was sustained for a salt deposit density (SDD) of 0.08 and for a non soluble contaminant deposit density (NSDD) of 0.48 mg/square cm.
Additional predictions include:
7. With SiR insulators or DC toughened glass, no added problems such as ion migration and thermal heating are expected under DC potential.
8. DC with impulse loading provides a higher air gap critical withstand level; impulses will be less onerous than equivalent AC.
9. The overhead earth-wires will help to attenuate and reduce the space charge field effects.
5.3 Proposal for Implementation
The national and regional grid diagrams are provided in appendix D. We have three proposals from this study.
I. Up rate one existing 275 kY circuit into KwaZulu Natal; starting in Northern Natal and terminating equally between Durban and Richards Bay: 400 km at 300 kY for 516 MW x 2 [1000 MW] power transfer.
2. Up rate one existing 400 kY circuit from Matimba Power Station to Midas Substation: 400 km at 400 kY for 1500 MW x 2 [3000 MW] power transfer.
3. Plan to up rate one existing 765 kY circuit from Alpha Substation to
Omega Substation: 1200 km at 800 kY for 4128 x 2 MW [8256 MW] power transfer; consider bi-directional values for optional power transfer directions.
This could form part of the next generation nuclear strategy; having the choice to move large bulk power in any direction depending on the contingency on hand.
For an estimate on expected costs of the proposal, we reference the last commissioned project in China.
The Three Gorges project has recently been commissioned in China. The first HYDC circuit built was a 500 kY, 3000 MW bipole. This was commissioned in 28 months at a cost of US$360m [8].
The 275 kY KwaZulu Natal conversion proposal is estimated to cost USD 200m; the 400 kY Matimba - Midas conversion proposal USD 400m and the 800 kY Alpha to Omega conversion proposal USD 600m. The cost data provided is a best guess and more accurate estimations would be required once the functional specifications are prepared.
The strength ofthe HYDC proposal is resident in the high current, high voltage capability of the power electronics. In addition, the reuse of existing assets to maximum capacity and the absence of additional servitudes need to be economically valued and added as net benefits emanating from the uprating exercise.
5.4 Recommendations for Further Study
Investigate and prepare IEC specifications for external insulation under DC potential. The work commenced by Vosloo et al [46] includes a comment on the need for a three times factor for the case of DC minimum specific creepage distance (mm/kV). This is attributed to the electrostatic catch and absence of zero crossings in the leakage currents for DC systems. However, service experience of composite insulators as reported in Cigre Electra publication No.16I [47]
indicates that no such correction factor is justified. Thus, more work, under IEC leadership, is required for DC external insulation standards.
Integrating HYDC into an existing HVAC interconnected power system reqUIres extensive planning and modelling for all cases of steady state stability, transient and dynamic stability.
At China Southern in Guangzhou, we were invited to explore the power system planning laboratory; a collection of real time digital simulators supported by PSCAD [48,49] calculating tools. The RTDS was shown to be the ideal tool for the power system studies; preparing both the technical proposals and also the commissioning parameters for the new incoming HYDe. In the time period between the planning proposals to that of real time commissioning, the real time system is continuously monitored for behaviours that follow normal power system faults with and without the HYDC in service. This is powerful modelling and adds great confidence to the planning proposals.
5.5 Conclusion
The key constraint of this research work was the lack of full scale testing facilities for high voltages of direct current. We visited all the South African test facilities and found no high voltage DC capability. We called upon international partners; Electric Power Research Institute of Lennox, New York, USA; State Grid of China Electric Power Research Institute; Swedish Test and Research Institute of Ludvika, Sweden and Power Grid of India, New Delhi. In all
cases we found either old facilities that were run down either in test equipment or measuring equipment or alternatively, new facilities are proposed and are currently under development.
We opted to do some work at the EPRI facilities in the USA but this has not materialised as much refurbishment work is required prior to testing. The option to employ the STRI laboratory in Ludvika was also considered; this high cost of contractual work formed part of the motivation to seek our own testing capability; to be located both at the University of Kwa-Zulu Natal HYDC Centre and at the National Electrical Test and Research Facility of the South African Bureau of Standards. Aside from the full scale testing capability, the current laboratory Cockcroft Walton HYDC Generator was employed to do small air gap experimentation.
Extrapolating the small air gap experimentation results to that of large air gaps as in full scale testing showed no correlation. Itwas best to stop the extrapolation process and rather evaluate the results as per the small air gap model. The interesting finding is that the laboratory based corona cage concept can be extended to form part of the transmission line tower configuration such that the corona energy could be captured and routed to the base of the tower and either employed directly or in association with other renewable energy sources as a potential energy source for on route power supplies. On route power supplies such as repeaters for telecommunications circuits, small rural lighting loads such as schools and clinics, safety lighting at river or rail or road crossings etc could be practically achieved. This work is recommended for the proposed 800 kV HYDC transmission that is being planned for Continental Africa. Further reading on the state of the art for tapping power from HY transmission lines is given by Nicolae et al [50].
Finally in the development of the new full scale laboratory facilities at the HYDC Centre, it is recommended that adequate test sources be made available for simultaneous bipolar test and measurement and for high current capacity with high voltage capability to continuously test and measure impact on external insulation. The continuous exposure of external insulation to high DC voltages is necessary to promote the electrostatic precipitation of atmospheric dirt in the vicinity of the electrode and the insulator assembly.
For the dissertation on hand, the lack of the full scale testing is considered not critical given the availability of multiple software modules for calculating and repeating the calculations for each parameter under study. These software tools were from different sources and their results
HVAC conditions. The calculations and results presented in both the appendix and the body of the report demonstrates the repeatability of the work done. In addition, under real operating conditions, the transmission lines experience a multitude of environmental variables and perform differently under the influence of the different variables. Thus both in service operating experience and full scale simulated laboratory results would form the ideal basis for further study and optimization of the design proposals. This solution would develop with time and experience. For the present day, the first pass study is adequate to commit a circuit or two for commercial operation; to start to gather the in field operating experience. A cautionary note for DC operations is to recommend that the transmission line be maintained employing live technology practices. In the case of the metallic earth return path; all the conductors in the bundle will carry induced currents and voltages and must be considered live at all times. In the case of insulator maintenance, full live technology tools and equipment must be employed at all times; maintaining the critical air gaps between conductor and tower body at all times. These air gaps will be under extreme electric field exposure and any disturbance either by bridging or by the addition of impurities could lead to air gap breakdown.